Detailed coseismic slip distribution of the 1944 Tonankai estimated from waveforms

by

Yuichiro TANIOKA1)

ABSTRACT Various instrumental data have been used to study the 1944 Tonankai earthquake. Coseismic slip distribution on the fault plane Kanamori [1972] used seismological data to of the 1944 Tonankai earthquake is estimated estimate the focal mechanism, and inferred from inversion of tsunami waveforms. The that the source area agreed with the one-day inversion result shows that a maximum slip aftershock distribution off the Kii Peninsula. of about 3 m occurred on the plate interface Recently, Kikuchi et al. [1999] estimated the off Shima peninsula. The total seismic seismic moment distribution using strong moment is estimated to be 2.0 X 1021 Nm (Mw motion waveforms recorded by the Japan 8.2). The result confirms that the 1944 Meteorological Agency. Ando [1975], Inouchi Tonankai earthquake did not rupture the and Sato [1975], and Ishibashi [1981] plate interface beneath the Tokai region and estimated the fault parameters using geodetic supports the existence of the seismic gap in data. A more detailed study based on geodetic the Tokai region. The slip of about 1.5 m on data [Sagiya and Thatcher, 1999] estimated the plate interface beneath Atsumi peninsula, the heterogeneous slip distribution on the northeast of the large slip area, is necessary down-dip side of the fault plane. However, the to explain the observed tsunami waveforms, geodetic data do not have resolution to although no seismic moment release was estimate the slips on the up-dip (offshore) estimated from strong motion data by side of the fault plane. In addition, the Kikuchi et al. [1999]. This may suggest that geodetic data cannot separate coseismic the rupture beneath Atsumi peninsula was deformation due to the 1944 Tonaknai slow. earthquake from that of the 1946 Nankai KEY WORDS: The 1944 Tonankai Earthquake earthquake, because the postseismic geodetic Slip Distribution survey was conducted after 1946. By using Tsunami waveforms tsunami waveforms, we can estimate the Tokai Earthquake heterogeneous slip distribution of the 1944 Tonankai earthquake independent to that of 1. INTRODUCTION the 1946 Nankai earthquake. Great interplate have repeatedly In this study, we perform tsunami occurred along the with an waveform inversion using a method similar to interval of about 120 years [Ando, 1975]. The ――――――――――――――――――― most recent events were the 1944 Tonankai 1) Senior Researcher, and and 1946 Nankai earthquakes. In this paper, Volcanology Department., Meteorological we concentrate on the 1944 Tonankai Research Institute, Tsukuba-shi, Ibaraki-ken, earthquake. 305-0052 , Japan that used by Satake [1993]. The estimated slip function for the inversion. The method of distribution is compared with the results tsunami waveform inversion is basically the obtained by other studies [Kanamori, 1972; same as that of Satake [1993]. The only Sagiya and Thatcher, 1999; Kikuchi et al., difference is that positivity constraints are 1999], and the differences and their assigned for the slip estimates. For error significance are discussed. In particular, analysis, the jackknife technique [see estimation of the detailed coseismic slip Tichelaar and Ruff, 1989] is applied. distribution in the northeastern end of the Nankai trough, the Tokai region, is important 3. SLIP DISTRIBUTIONS because the next large event is anticipated Large slip (> 2 m) is estimated on three [Ishibashi, 1981]. The Japanese government subfaults near Shima peninsula. The largest started an extensive earthquake prediction slip was estimated as about 3m off Shima experiment in the Tokai region [Mogi, 1981], peninsula. This large slip corresponds to the and various instruments have been installed source region estimated from the one-day [Yokota and Yamamoto, 1989]. aftershock distribution by Kanamori [1972]. The inversion indicates that there was no slip 2. DATA AND METHOD was on subfaults beneath the Tokai region. The fault located on the estimated upper This is consistent with the result of Ishibashi surface of the slab is divided into 23 subfaults [1981], suggesting that the rupture of the by varying the depth and dip angle. The 1944 Tonankai earthquake did not extend into subfault size is 45 km X 45 km, the strike is the Tokai area which was previously ruptured 240°, and the rake angle is 110° for all the by the 1854 Tonankai earthquake. The total subfaults. The strike and rake angles are the seismic moment is calculated as 2.0 X 1021 Nm same as Satake [1993]. Tsunami waveform assuming that the rigidity is 5 X 1010 N/m2. data recorded at the 10 tide gauge stations are used. The tsunami waveform for each tide 4. DISCUSION gauge station consists of 40 to 110 min of data The large (> 2 m) slip region is consistent with a sampling interval of 1 min. The with the large moment release region tsunami is numerically computed on actual estimated from strong motion seismograms by bathymetry. The finite-difference Kikuchi et al. [1999], and also consistent with computations for the linear long-wave the large slip region near Shima peninsula equations are carried out. The grid size is estimated from geodetic data by Sagiya and basically 20 sec of arc (about 600 m), but finer Thatcher [1999]. In general, the slip and grids (4 sec) are used. The time step of the moment distribution patterns estimated for computation is 1.5 s to satisfy a stability the 1944 Tonankai earthquake from three condition of all grid systems. The initial data sets, seismic waves, tsunami, and condition of tsunami propagation is an ocean geodetic data sets, are similar although the bottom deformation, which is computed using seismic moments estimated from each data the equations of Okada [1985]. Tsunami set have some differences. In detail, there is a waveforms at the tide gauge stations are discrepancy near the Atsumi peninsula where computed from each subfault with a unit our estimated slips on two subfaults are 1.5 m amount of slip, and used as the Green’s and 1.7 m, respectively. Using strong motion data, Kikuchi et al. [1999] estimated no seismic moment release beneath the Atsumi 2. Inouchi, N, and H. Sato, Vertical crustal peninsula, while Sagiya and Thatcher [1999] deformation accompanied with the Tonankai estimated slip of 1.2 m, similar to our result. earthquake of 1944, Bull. Geogr. Surv. Inst. Inouchi and Sato [1975] and Ishibashi [1981] Jpn., 21, 10-18, 1975. also indicated that slip beneath the Atsumi peninsula is necessary to explain the geodetic 3. Ishibashi, K., Specification of a soon to data. The discrepancy may suggest that a occur seismic faulting in the Tokai district, slow rupture of the plate interface beneath central Japan, based upon seismotectonics, in the Atsumi peninsula excited the tsunami but Earthquake prediction: An International not short period seismic waves. Review, Maurice Ewing Ser., vol. 4, edited by D.W. Simpson and P.G. Richards, pp. 297-332, 5. CONCLUSION AGU, Washington, D.C., 1981. The slip distribution of the 1944 Tonankai earthquake estimated from tsunami 3. Kanamori, H., Tectonic implications of the waveforms shows that a maximum slip of 3.3 1944 Tonankai and the 1946 Nankaido m occurred on the plate interface off Shima earthquakes, Phys. Earth Planet Inter., 5, peninsula. The large slip is consistent with 129-139, 1972. other results obtained from strong motion data by Kikuchi et al. [1999] and from 4. Kikuchi, M., M. Nakamura, M. Yamada, geodetic data by Sagiya and Thatcher [1999]. and K. Yoshikawa, Source process of the 1944 Slip of about 1.5 m on the plate interface Tonankai earthquake inferred from JMA beneath the Atsumi peninsula is necessary to strong motion records, Programme and explain the tsunami waveforms. Abstracts, The seismological society of Japan The result confirms that the 1944 Tonankai Fall Meeting, A57, 1999. earthquake did not rupture the plate interface in the northeastern end of the 5. Mogi, K., Earthquake prediction program Nankai trough beneath the Tokai region in Japan, in Earthquake prediction: An which was ruptured by the penultimate International Review, Maurice Ewing Ser., vol. earthquake in 1854. The result supports the 4, edited by D.W. Simpson and P.G. Richards, existence of seismic gap suggested by pp. 635-666, AGU, Washington, D.C., 1981. Ishibashi [1981]. Therefore, continuation of the extensive earthquake prediction 6. Okada, Y., Surface deformation due to experiments in the Tokai region conducted by shear and tensile faults in a half-space, Bull. the Japanese governments, such as the Japan Seismol. Soc. Am., 75, 1135-1154, 1985. Meteorological Agency [Yokota and Yamamoto, 1989], is important to the hazard mitigation. 7. Sagiya, T. and W. Thatcher, Coseismic slip resolution along a plate boundary REFERENCES megathrust: The Nankai Trough, southwest 1. Ando, M., Source mechanisms and tectonic Japan, J. Geophys. Res., 104, 1111-1129, 1999. significance of historical earthquakes along the Nankai trough, Japan, Tectonophysics, 27, 8. Satake, K., Depth distribution of coseismic 119-140, 1975. slip along the Nankai Trough, Japan, from joint inversion of geodetic and tsunami data, J. Geophys. Res., 98, 4553-4565, 1993.

9. Tanioka, Y and K. Satake, Coseismic slip distribution of the 1946 Nankai earthquake and aseismic slips caused by the earthquake, Earth Planets Space, 53, 235-241,2001.

10. Tichelaar, B. W., and L. J. Ruff, How good are our best model? Jackknifing, bootstrapping, and earthquake depth, Eos, 70, 593, 605-606, 1989.

11. Yokota, T., and M. Yamamoto, On the Earthquake Phenomena Observation System (Outline of the system) (in Japanese), Quarterly J. Seismology, 52, 89-99, 1989.